204 CEYLON PEARL OYSTER REPORT. 



Riicksicht anf ihre Basis als Species einer neuen Gattuug zu beschreiben, wobei ich es 

 Anderen ttberlasseu muss, die schon beschriebenen lebenden und fossilen Eupsammidse 

 nach den durch die philippinischen Form en sicb ergebenden Andeutungen zu unter- 

 suchen, und mit diesen in systematischen Zusainmenhang zu stellen." Notwith- 

 standing Skmper's hope that one of his successors would show the proper systematic 

 relation between his Philippine specimens and previously described fossil and living 

 forms, Duncan, in his 'Revision of the Madreporaria,' retains Semper's genus 

 Rhodopsammia, but removes it from the " alliance' Balanophylloida, to which, on its 

 author's own testimony, it belongs, and places it apart as a genus which cannot be 

 included in any alliance ! 



The Eupsammiidse in Professor Herdman's collection include free and fixed, solitary 

 and colonial forms, and most of them resemble one another very closely in such 

 distinctive characters as septal arrangement, columella, and costse. The solitary 

 forms (with the exception of Heteropsammia) must, without doubt, be referred to a 

 single genus, and they agree in all respects with Semper's definition of the genus 

 Rhodopsammia. But if we agree with Semper and ignore, as we must after the 

 study of a sufficient number of specimens, such variable characters as attachment or 

 its contrary, the presence or absence or the completeness or incompleteness of a fifth 

 cycle of septa, or the relative thickness of the septa, we must recognise that the 

 genus Rhodopsammia, Semper, includes Eupsammia, M. Edwards and Haime ; 

 Balanophyllia, Searles Wood ; Leptopsammia, M. Edw. and Haime ; Endo- 

 psammia, M. Edw. and Haime, and it becomes a question as to what name shall be 

 used to denote the single genus into which all the other genera are absorbed. The 

 genus Balanophyllia was founded by Searles Wood in 1844, and there is no 

 objection to be taken to it on the score of iudetiniteness. Therefore, by the rules of 

 nomenclature, it has the priority over the genera founded by M. Edwards and 

 Haime in 1848, and a fortiori over Semper's genus Rhodopsammia founded in 1872. 

 Hence I suggest the amendment of the definition of Balanophyllia in the terms of 

 Semper's definition of Rhodopsammia, the latter name being dropped and 

 Eupsammia, Leptopsammia, and Endopsammia merged into Balanophyllia. 



It is a characteristic, though not a peculiarity of the Eupsannniidie, that the septa 

 of the later cycles curve towards and are usually united to those of the preceding 

 cycle, and it commonly happens that the septa of what on the ordinary system of 

 reckoning would be called the last cycle are larger than those of the preceding cycle. 

 Semper, many years ago (49, p. 25y), called attention to the inapplicability of 

 M. Edwards and Haime's law of septal sequence to the Eupsammiidpe as well as to 

 other Madreporaria described in his well-known memoir, but until recently there has 

 been no satisfactory explanation of the apparent irregularity in the septal sequence 

 of these forms, and it has been necessary, in spite of its obvious disadvantages, to 

 retain M. Edwards and Haime's system of notation in describing the septa of all 

 Madreporaria. But as long ago as 1871, Pourtales (43) gave a description of the 



